Chronic fatigue syndrome (CFS) is a complex multi-system disorder, which is often misdiagnosed as a psychiatric illness. As a result, the diagnosis of CFS is highly controversial. Discovery of CFS-specific biomarkers that can differentiate the disorder from phenotypically similar psychiatric conditions, such as major depressive disorder (MDD), could thus have a profound impact, not only for how the disorder is generally perceived and managed, but also for the development of objective diagnostic tests, for identification of new therapeutic targets, as well as for advancing scientific understanding of CFS. Recently, using advanced magnetic resonance neuroimaging techniques and a standardized battery of clinical assessments in 15 patients with CFS with those in 15 patients with MDD and in 13 healthy controls, the applicants discovered strong experimental evidence, including a mean deficit of 36% in the most abundant antioxidant in living tissue, glutathione (GSH), increased ventricular cerebrospinal fluid (CSF) lactate, and decreased regional cerebral blood flow (rCBF) compared to controls, which suggested increased oxidative stress as a pathophysiological model of CFS. However, while highly promising and intrinsically consistent, both the validity and the specificity of this oxidative stress hypothesis for CFS remain uncertain, as (a) the essential findings of the study have yet to be replicated, and (b) the same types of abnormalities were found in MDD compared to controls. On the other hand, with comparisons revealing trend-level differences between CFS and MDD, the investigators hypothesized that limited sample size, coupled with the inherent clinical heterogeneity of the two disorders, likely limited the power of their pilot study to detect potential differences between the two disorders. Therefore, to address this potential limitation and to attempt objective differentiation of CFS and MDD - a daunting and continuing challenge - the investigators propose: (1) to replicate in larger cohorts the results of their pilot neuroimaging study that suggested the oxidative stress hypothesis of CFS; (2) to extend the support and evidence base for the model through measurements of several established markers of oxidative stress in plasma, urine and CSF samples from all the subjects; (3) to correlate the resulting objective outcome measures with clinical indices of overall health and functional disability in all subjects; and (4) to attempt to decrease the inherent clinical heterogeneity in both the CFS and MDD groups through stratification or subtyping techniques based on clinical variables that are unique to each disorder, and then to compare the outcome measures between the resulting subgroups. The expectation is that this approach would identify subgroups of CFS and MDD patients between which significant differences in outcome measures exist that can enable objective differentiation of the two disorders, thereby establishing the outcome measures as bona fide diseases biomarkers, and supporting oxidative stress as a valid and specific pathophysiological model for CFS.